• State Resolved
  • Locked Locked
  • Replies 20 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 1345 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

PGA870: Need design review and confirmation on PGA870 + ADS5560

Arun Kumar4
Prodigy 60 points
Part Number: PGA870
Other Parts Discussed in Thread: ADS5560, TINA-TI

Hi TI team,

First of all, thank you very much for considering our tech support request.

We are designing an analog front-end circuit including PGA870 and ADS5560 for acquiring the analog signal output from the photovoltaic detector.

The current output from the photovoltaic detector is converted to desired voltage using a preamplifier with a trans-impedance opamp.

The output from the amplifier is provided to PGA870 and then to ADS5560 ADC. The schematics of the same are attached for your review.

Kindly let us know whether our circuit will be sufficient and suitable for our application. Also, let us know if we needed to add any other circuits.

Reference Ticket No: CS0792702

  • 0
    TI__Mastermind 28865 points

    Hello Arun,

      The schematic looks good. Quick question on the input filtering for the ADC, what is the frequency range and characteristics of your signal? At a glance, it looks normal amount of values for charge bucket. But, I can simulate this on our end, and make sure stability wise and if it will settle for this ADC.

    Thank you,

    Sima 

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,

    Thank you for your quick response.

    The laser beam will be emitted from a laser source as a repetitive pulse at a frequency of 20 Hz. The reflected laser pulse at the same frequency will be received by the photodetector and provided to the PGA via trans-impedance opamp-based preamplifier circuit.

    Regards,
    Arun Kumar V

  • 0 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hello Arun,

      Just to clarify, I believe you provided 1/time interval between pulses, but what is the rise time of the pulse? This is important due to also the AC coupling at the input and output of the PGA. It will cutoff frequencies below 10kHz. But, with the previous stage being a TIA based design, it is highly likely you are looking in the few ns rise time? Still need to look in-depth at the filtering at input of the ADC + settling; I will get back to you on this by early next week. Sorry for the delay on second part. 

    Thank you,
    Sima 

  • 0 in reply to Sima Jalaleddine
    TI__Mastermind 28865 points

    Hello Arun, 

      Are there any updates on the pulse bandwidth range? I have attached the simulation on the amplifier + filtering of your design below. Let us know if there are any further questions. 

    PGA870.tsc

    Thank you,

    Sima

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,


    Sorry for the delay.

    We have changed our input and output filter sections in the circuit with respect to our requirements.

    An example of the input pulse (detector signal) is attached for your reference. The on period of the pulse will be in sub 5 nano seconds and the frequency will be 20Hz.

    The simulated schematics file (.tsc) of the updated circuit is also attached for your reference. Please review the circuit and provide your comments on the same.

    PGA870TI.TSC

  • +1 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hello Arun,

      Thank you for the additional comments and providing a Tina-TI file. I will review the circuit, and provide feedback by the end of the week.

    Best Regards,

    Sima

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,

            Any updates on our query?

    Thanks& Regards,

    Arun kumar

  • 0 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hi Arun,

      Yes, I reviewed your circuit. Thank you for using Tina-TI to show your updates. 

      That is a good idea to place resistors prior to your capacitances at the output. The circuit is stable (no peaking, ringing, and output noise response looks good), but now with an increased passband in the frequency response. You could probably cut this down from around 500MHz to 250-300MHz by increasing the loading cap (tuning the output stage.

    Thank you,

    Sima

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,

    Thank you for your kind response. 

    Can we have a transformer coupling at the input section rather then having high value capacitor of 220uF (Electrolytic)? Please share your comments.

    Thank you,

    Arun Kumar

  • 0 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hello Arun,

       You could, but your bandwidth is around 20Hz, which is very low. It might be hard to find a transformer that can accept frequencies that low. You might not even need AC coupling at these frequencies. Most likely would need to focus on filtering out higher frequency noise instead. 

       Is the purpose of these high valued capacitors to remove a DC offset? 

    Thank you,

    Sima 

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,

    Ours signal is single ended, hence we need a AC coupling at the input stage. 

    We have a very high value capacitor in the input which we simulated and confirmed. We doubt on having this big capacitor for analog singles which will typically a electrolytic.

    Can you suggest us an optimal value capacitor for the input (220uF) and output (330uF) by simulating it?

    Thanks,

    Arun

  • 0 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hello Arun,

      You can dc couple our FDAs even as single-ended source. The only issue is if there is a DC signal that you would need to eliminate with these large capacitors. 

    Edit: Do not drive Vmid pins as earlier suggested, bypass + grounding of these pins is correct. 

    Thank you,

    Sima

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,

    Sorry I missed to mention our source voltage which can range between +/- 1V. Hence we need AC coupling at the input stage.

    Thank you,

    Arun

  • 0 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hello Arun,

      Sorry one more additional clarification. That would mean that the TIA preamplifier has a +/-1V DC offset at the output of voltage pulse with a rise time of around 15ms? 

    Thank you,

    Sima 

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,

    We have attached the specification of the pre-amplifier. The output voltage swing is +/-1V. The rise time will be in <10nsec is what we understand from our customer.

    Regards,

    Arun V

  • +1 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hello Arun,

       Thanks, this information helps a lot! That would mean you will be inputting an +/-1V <10nsec pulse with an offset voltage of around +/-10mV or +/-25mV. This means you would need greater than 35MHz overall system bandwidth, and AC coupling wouldn't be a critical issue due to 25mV < than 200mV specified in PGA's datasheet. 

       However, there are other issues that might lead to needing to change PGAs. Is the +/-1V a fixed output voltage? If so, you will run into issues with output voltage range depending on your gain settings: 

        Is there a reason you need to use a PGA? Or is the +/-1V output voltage range swing the range of the device, and the application's output voltage might fall within this range depending on light intensity or photodiode connection? If not, it might be more beneficial to use a fully differential amplifier with a VCM that can be set externally via ADC. 

    Thank you,

    Sima 

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,

    Thank you for your kind response.

    The +/-1V output of the preamplifier is not a fixed voltage. It will vary from +/-200mV to +/-1V based on the laser pulse received.
    Hence to amplify the lower voltage of +/- 200mV, we need a PGA.
    Our only concern is having the bulk capacitors at the input (220uf) and output (330uF) which we simulated and arrived. Please suggest us an optimal value for those capacitors for our requirement.

    Regards
    Arun V

  • +1 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hello Arun,

      Sorry for the delay in response. 

      Yes, then the PGA will work. But, do take caution regarding the output common range of the PGA vs gain selection.

      For the AC coupling, since we are looking at 35MHz design, you would actually be able to decrease these bulk capacitors for 0.01uF at around 300kHz to 1uF at around 5kHz. It would be best to switch over to ceramic capacitors due to their low ESL and ESR, and wouldn't have to worry about polarity. 

    6683.PGA870TI.TSC

    Thank you,

    Sima 

  • 0 in reply to Sima Jalaleddine
    Prodigy 60 points

    Hi Sima,

    Thanks for your suggestion.

    we have simulated the circuit with different capacitors at the input and output. We are not getting the expected results.

    The table attached shows the observed readings at the input and output. Also we have attached the image for your reference

    Input Input capacitors Output Capacitors Input Voltage Observed Output Voltage Observed
    Sine wave-100mV @20Hz 10nF 10nF 10uV 10nV
    Sine wave-100mV @20Hz 1uF 1uF 1mV 100uV


    Looking forward for your inputs.

    Regards,
    Arun

  • +1 in reply to Arun Kumar4
    TI__Mastermind 28865 points

    Hello Arun,

      Thank you for trying out the suggested values. The reason is due to these capacitors attenuating the signal at 20Hz input due to the suggested AC coupling cap values. However, with a pulse rise time of <10ns, the bandwidth you would need is around 0.35/10ns which is around 35MHz rather than 20Hz for your application. Or will the input to this PGA be a 20Hz sin wave? Sin waves have different relationships to bandwidth compared to pulse/square waves coming from a TIA.

      Below is an example of pulse at 100mV with rise time of 10ns. AC response shows 13ish dB of gain at the output after attenuation.

    6082.PGA870TI.TSC

    Thank you,
    Sima